
# Importing libraries
import pandas as pd
import datetime as dt
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px
import plotly.io as pio

# Setting aesthetic parameters for plots
plt.rcParams['figure.figsize'] = (16, 8)
pio.templates.default = 'presentation'
sns.set_style('whitegrid')
# Turning off scientific notation in displaying numbers
pd.set_option('display.float_format', '{:.5f}'.format)
# Turning off warnings for chained assignments
pd.options.mode.chained_assignment = None


# Importing data and declaring variables corresponding to different quarters
q2_2019 = pd.read_csv('divvy_data/Divvy_Trips_2019_Q2.csv')
q3_2019 = pd.read_csv('divvy_data/Divvy_Trips_2019_Q3.csv')
q4_2019 = pd.read_csv('divvy_data/Divvy_Trips_2019_Q4.csv')
q1_2020 = pd.read_csv('divvy_data/Divvy_Trips_2020_Q1.csv')


# List comprehension for printing the column names of each dataframe
[data.columns for data in [q2_2019, q3_2019, q4_2019, q1_2020]]

[Index(['01 - Rental Details Rental ID', '01 - Rental Details Local Start Time',
        '01 - Rental Details Local End Time', '01 - Rental Details Bike ID',
        '01 - Rental Details Duration In Seconds Uncapped',
        '03 - Rental Start Station ID', '03 - Rental Start Station Name',
        '02 - Rental End Station ID', '02 - Rental End Station Name',
        'User Type', 'Member Gender',
        '05 - Member Details Member Birthday Year'],
       dtype='object'),
 Index(['trip_id', 'start_time', 'end_time', 'bikeid', 'tripduration',
        'from_station_id', 'from_station_name', 'to_station_id',
        'to_station_name', 'usertype', 'gender', 'birthyear'],
       dtype='object'),
 Index(['trip_id', 'start_time', 'end_time', 'bikeid', 'tripduration',
        'from_station_id', 'from_station_name', 'to_station_id',
        'to_station_name', 'usertype', 'gender', 'birthyear'],
       dtype='object'),
 Index(['ride_id', 'rideable_type', 'started_at', 'ended_at',
        'start_station_name', 'start_station_id', 'end_station_name',
        'end_station_id', 'start_lat', 'start_lng', 'end_lat', 'end_lng',
        'member_casual'],
       dtype='object')]


# Looking at a sample of the q2_2019 dataset
q2_2019.sample(2)


# Renaming the q2_2019 columns in accordance with the 2020 standard
q2_2019 = \
q2_2019.rename(columns={
    '01 - Rental Details Rental ID': 'ride_id',
    '01 - Rental Details Local Start Time': 'started_at',
    '01 - Rental Details Local End Time': 'ended_at',
    '03 - Rental Start Station ID': 'start_station_id',
    '03 - Rental Start Station Name': 'start_station_name',
    '02 - Rental End Station ID': 'end_station_id',
    '02 - Rental End Station Name': 'end_station_name',
    'User Type': 'member_casual'
})


# Rewriting the dataframe so as to only include columns that also
# exist in the q1_2020 dataframe
q2_2019 = \
q2_2019.loc[:, q2_2019.columns.isin(q1_2020.columns)]
# Validating the result
q2_2019.sample(5)


# Renaming the q3_2019 columns in accordance with the 2020 standard
q3_2019 = \
q3_2019.rename(columns={
    'trip_id': 'ride_id',
    'start_time': 'started_at',
    'end_time': 'ended_at',
    'from_station_id': 'start_station_id',
    'from_station_name': 'start_station_name',
    'to_station_id': 'end_station_id',
    'to_station_name': 'end_station_name',
    'usertype': 'member_casual'
})


# Rewriting the dataframe so as to only include columns that also
# exist in the q1_2020 dataframe
q3_2019 = \
q3_2019.loc[:, q3_2019.columns.isin(q1_2020.columns)]
# Validating the result
q3_2019.sample(5)


# Renaming the q4_2019 columns in accordance with the 2020 standard
q4_2019 = \
q4_2019.rename(columns={
    'trip_id': 'ride_id',
    'start_time': 'started_at',
    'end_time': 'ended_at',
    'from_station_id': 'start_station_id',
    'from_station_name': 'start_station_name',
    'to_station_id': 'end_station_id',
    'to_station_name': 'end_station_name',
    'usertype': 'member_casual'
})


# Rewriting the dataframe so as to only include columns that also
# exist in the q1_2020 dataframe
q4_2019 = \
q4_2019.loc[:, q4_2019.columns.isin(q1_2020.columns)]
# Validating the result
q4_2019.sample(5)


# Now doing the same for the 2020 dataset to exclude data
# not present in previous datasets
q1_2020 = \
q1_2020.loc[:, q1_2020.columns.isin(q4_2019.columns)]
# Validating the result
q1_2020.sample(5)


# Changing the order of the columns in q1_2020 to conform to the rest
q1_2020 = \
q1_2020.loc[:, q4_2019.columns]
# Validating the result
q1_2020.sample(5)


# The all() function returns True only if all boolean values
# in the array are True
all(q2_2019.columns == q3_2019.columns) \
and all(q3_2019.columns == q4_2019.columns) \
and all(q4_2019.columns == q1_2020.columns)

True


# Concatenating data into a single dataframe, df
df = pd.concat([q2_2019, q3_2019, q4_2019, q1_2020])
# Validating the result
df.sample(5)


# Printing dataframe's general information
df.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 3879822 entries, 0 to 426886
Data columns (total 8 columns):
 #   Column              Dtype  
---  ------              -----  
 0   ride_id             object 
 1   started_at          object 
 2   ended_at            object 
 3   start_station_id    int64  
 4   start_station_name  object 
 5   end_station_id      float64
 6   end_station_name    object 
 7   member_casual       object 
dtypes: float64(1), int64(1), object(6)
memory usage: 266.4+ MB


# Counting the total amount of duplicates
df.duplicated().sum()

0


# Showing the amount of missing values by column
df.isna().sum()

ride_id               0
started_at            0
ended_at              0
start_station_id      0
start_station_name    0
end_station_id        1
end_station_name      1
member_casual         0
dtype: int64


# Seeing the rows with the missing values
df.query('end_station_id.isna() \
         or end_station_name.isna()')


# Deleting rows with missing values
df = df[df.end_station_id.notnull()]
# Validating the result
df.isna().sum()

ride_id               0
started_at            0
ended_at              0
start_station_id      0
start_station_name    0
end_station_id        0
end_station_name      0
member_casual         0
dtype: int64


# Printing column data types
df.dtypes

ride_id                object
started_at             object
ended_at               object
start_station_id        int64
start_station_name     object
end_station_id        float64
end_station_name       object
member_casual          object
dtype: object


# Formatting time and date values from string to datetime
df.started_at = pd.to_datetime(df.started_at, format='%Y-%m-%d %H:%M:%S')
df.ended_at = pd.to_datetime(df.ended_at, format='%Y-%m-%d %H:%M:%S')


df['end_station_id'].sample(5)

1434600    94.00000
729622    268.00000
135345     40.00000
863457    420.00000
120870    174.00000
Name: end_station_id, dtype: float64


# Converting the column from float to integer numbers
df['end_station_id'] = df['end_station_id'].astype('int')


# Renaming the 'member_casual' column
df = df.rename(columns={
    'member_casual': 'is_member'
})


# Seeing different values in the 'is_member' column
df['is_member'].value_counts()

Subscriber    2595461
Customer       857474
member         378407
casual          48479
Name: is_member, dtype: int64


# Standardizing different member/non-member classifications as 1 and 0
df.loc[df['is_member'].isin(['Subscriber', 'member']), 'is_member'] = 1
df.loc[df['is_member'].isin(['Customer', 'casual']), 'is_member'] = 0


# Transforming the column to boolean type
df['is_member'] = df['is_member'].astype('bool')
# Validating the transformation
df['is_member'].value_counts(normalize=True)

True    0.76650
False   0.23350
Name: is_member, dtype: float64


# Declaring the duration column as the difference 
# between the columns ended_at and started_at
df['duration'] = df['ended_at'] - df['started_at']


# Extracting the total number of seconds from the datetime data type
df['duration'] = df['duration'].dt.total_seconds().astype(int)
# Eenaming the column to indicate the unit
df = df.rename(columns={
    'duration': 'duration_sec'
              })
# Validating the result
df.duration_sec.sample(5)

663073      481
1303879     338
704609      785
702502      673
769829     1446
Name: duration_sec, dtype: int64


# Extracring and rounding the total number of minutes from the
# Duration_sec column
df['duration_min'] = round(df['duration_sec']/60).astype('int')


# Extracting the weekday number from the started_at column
# Where 0 is Monday and 6 is Sunday
df['weekday'] = pd.DatetimeIndex(df['started_at']).weekday
# Validating the result
df.weekday.sample(5)

633112     6
225632     3
103038     5
1470869    3
1030407    2
Name: weekday, dtype: int64


# Extracting the month number where 1 is January and 12 is December
df['month'] = pd.DatetimeIndex(df['started_at']).month
# Validating the result
df.month.sample(5)

656172      6
1321978     9
321037     10
571457      5
393976      3
Name: month, dtype: int64


# Extracting the hour from the started_at column
df['hour'] = pd.DatetimeIndex(df['started_at']).hour
# Validating the result
df.hour.sample(5)

357787    12
362808    19
596405    19
192050     7
501024    11
Name: hour, dtype: int64


print(f'''Maximum ride duration: \
{round(df["duration_min"].max()/60)} hours,
Minimum ride duration: \
{df["duration_min"].min()} min''')

Maximum ride duration: 2608 hours,
Minimum ride duration: -56 min


# Sorting ride durations to find percentiles of chosen cutoff values
duration_sorted = df['duration_sec'].sort_values()


# Finding the percentiles of chosen cutoff values by 
# dividing the index of their position in the sorted Series, 
# dividing it by the length of the Series, 
# multiplying by 100 and rounding to the second decimal
print(f'''40 seconds is in the \
{round(duration_sorted.searchsorted(40)
/len(duration_sorted)*100, 2)}%-th percentile of values in the\
 duration_sec column.
8 hours is in the \
{round(duration_sorted.searchsorted(8*60*60)
/len(duration_sorted)*100, 2)}%-th percentile.''')

40 seconds is in the 0.18%-th percentile of values in the duration_sec column.
8 hours is in the 99.86%-th percentile.


# Creating a dataframe without rows with implausible ride durations 
df_cleaned = \
df[
    (df['duration_sec'] >= 40)
    & (df['duration_sec'] <= 8*60*60)
]


# Finding the share of rows deleted from the dataset, rounding to the third decimal 
# and expressing as a percentage
print(f'''\
trimming {round((df.shape[0] - df_cleaned.shape[0])
/df.shape[0], 3)*100}\
% of data''')

trimming 0.3% of data


# Overwriting the original dataframe variable
# with the cleaned data
df = df_cleaned


# Computing the number of unique station id's
print(f'Number of unique stations: {df.start_station_id.nunique()}')

Number of unique stations: 618


# Checking if all ride id's are unique
df.shape[0] == df.ride_id.nunique()

True


# Creating a pivot table for visualization
df_pivot_months = (
    pd.pivot_table(
        # Grouping by month and membership status
        df,
        index=['month', 'is_member'],
        values='ride_id',
        # Counting the number of rides
        aggfunc={
            'ride_id': 'count'
        }).reset_index().rename({'ride_id': 'count'}, axis=1)
    # Replacing boolean values for visualization
    # purposes
    .replace({
        False: 'casual riders',
        True: 'members'
    }))
df_pivot_months['month'] = \
df_pivot_months['month'].replace(
    # Replacing integers with months for better visualization
    {1:'Jan',
     2:'Feb',
     3:'Mar',
     4:'Apr',
     5:'May',
     6:'Jun',
     7:'Jul',
     8:'Aug',
     9:'Sep',
     10:'Oct',
     11:'Nov',
     12:'Dec'})
# Validating the pivot table
df_pivot_months.head(5)


# Plotting differences across months
# with the plotly express library
fig = px.bar(
    df_pivot_months,
    x='month',
    y='count',
    color='is_member',
    # Changing labels for clarity
    labels={
        'count': 'Number of rides',
        'month': 'Month',
        'is_member': ''
    },
    # Unstacking member and
    # casual rider bars
    barmode='group',
    # Formatting the count number output
    text_auto='.3s',
    # Setting the title
    title="Number of rides per month")
# Adjusting count number traces
fig.update_traces(textfont_size=13, textposition="outside")
# Adjusting the grid
fig.update_yaxes(gridwidth=2)
fig.show()


# Creating a pivot table for visualization
df_pivot_weekdays = (
    pd.pivot_table(
        df,
        # Grouping by weekday and membership status
        index=['weekday', 'is_member'],
        values='ride_id',
        # Counting the number of rides
        aggfunc='count').reset_index()
    # Replacing boolean values for visualization
    # purposes
    .replace({
        False: 'casual riders',
        True: 'members'
    }).rename({'ride_id': 'count'}, axis=1))
df_pivot_weekdays = df_pivot_weekdays.replace({
    # Replacing integers with weekdays for better visualization
    0: 'Mon',
    1: 'Tue',
    2: 'Wed',
    3: 'Thu',
    4: 'Fri',
    5: 'Sat',
    6: 'Sun'
})
# Validating the pivot table
df_pivot_weekdays.head(5)


# Plotting differences across weekdays
fig = px.bar(
    df_pivot_weekdays,
    x='weekday',
    y='count',
    color='is_member',
    # Changing labels for clarity
    labels={
        'count': 'Number of rides',
        'weekday': 'Weekday',
        'is_member': ''
    },
    # Unstacking member and
    # casual rider bars
    barmode='group',
    # Formatting the count number output
    text_auto='.3s',
    title="Number of rides per weekday")
# Adjusting count number traces
fig.update_traces(textfont_size=13, textposition="outside")
# Adjusting the grid
fig.update_yaxes(gridwidth=2)
fig.show()


# Creating a pivot table for visualization
df_pivot_hours_workdays = (
    pd.pivot_table(
        # Choosing only weekdays from 0 to 4 (Mon-Fri)
        df[df['weekday'] <= 4],
        # Grouping by hour and membership status
        index=['hour', 'is_member'],
        values=['ride_id', 'duration_min'],
        # Counting the number of rides
        # as well as finding the median duration
        aggfunc=({
            'ride_id': 'count',
            'duration_min': 'median'
        })).reset_index()
    # Replacing boolean values for visualization
    # purposes
    .replace({
        False: 'casual riders',
        True: 'members'
    }).rename({'ride_id': 'count'}, axis=1))
# Validating the pivot table
df_pivot_hours_workdays.head(5)


# Plotting differences across hours
fig = px.bar(
    df_pivot_hours_workdays,
    x='hour',
    y='count',
    color='is_member',
    # Including hover data with ride duration
    hover_data=['duration_min'],
    # Changing labels for clarity
    labels={
        'count': 'Number of rides',
        'hour': 'Hour',
        'is_member': '',
        'duration_min': 'Median duration'
    },
    # Unstacking member and
    # Casual rider bars
    barmode='group',
    # Adjusting the title
    title='Number of rides per hour on workdays')
# Showing all of the x axis labels
fig.layout.xaxis.dtick = 1
# Adjusting the grid
fig.update_yaxes(gridwidth=2)
fig.show()


# Creating a pivot table for visualization
df_pivot_hours_weekends = (
    pd.pivot_table(
        # Choosing only weekdays 5 and 6 (Sat, Sun)
        df[df['weekday'] >= 5],
        # Grouping by hour and membership status
        index=['hour', 'is_member'],
        values='ride_id',
        # Counting the number of rides
        aggfunc='count').reset_index()
    # Replacing boolean values for visualization
    # Purposes
    .replace({
        False: 'casual riders',
        True: 'members'
    }).rename({'ride_id': 'count'}, axis=1))
# Validating the pivot table
df_pivot_hours_weekends.head(5)


# Plotting differences across hours
fig = px.bar(
    df_pivot_hours_weekends,
    x='hour',
    y='count',
    color='is_member',
    # Changing labels for clarity
    labels={
        'count': 'Number of rides',
        'hour': 'Hour',
        'is_member': ''
    },
    # Unstacking member and
    # casual rider bars
    barmode='group',
    title="Number of rides per hour on weekends")
# Showing all of the x axis labels
fig.layout.xaxis.dtick = 1
# Adjusting the grid
fig.update_yaxes(gridwidth=2)
fig.show()


# plotting differences in ride durations
# using seaborn and not plotly due to
# plotly crashing when processing the data
fig = sns.violinplot(data=df,
                     x=df['is_member']\
                     # replacing boolean values for visualization
                     # purposes
                     .replace({False:'casual riders',
                               True:'members'}),
                     y='duration_min',
                     # standardizing the visual style
                     saturation=0.9)
fig.set_title('Ride duration by membership status')
fig.set_xlabel('')
fig.set_ylabel('Duration in minutes');


# plotting differences in ride durations
fig = sns.violinplot(data=df,
                     x=df['is_member']\
                     # replacing boolean values for visualization
                     # purposes
                     .replace({False:'casual riders',
                               True:'members'}),
                     y='duration_min',
                     # standardizing the visual style
                     saturation=0.9)
fig.set_title('Ride duration by membership status')
fig.set_xlabel('')
fig.set_ylabel('Duration in minutes')
# limiting the y axis
plt.ylim(0, 100)

# embedding calculations for menas and medians inside f-strings
print(f'''The median and mean durations of rides for members are \
{round(df.loc[df['is_member'] == 1, 'duration_min'].median())}min and \
{round(df.loc[df['is_member'] == 1, 'duration_min'].mean())}min.
For non-members the statistics are \
{round(df.loc[df['is_member'] == 0, 'duration_min'].median())}min and \
{round(df.loc[df['is_member'] == 0, 'duration_min'].mean())}min,\
respectively.''')

The median and mean durations of rides for members are 10min and 13min.
For non-members the statistics are 26min and 37min,respectively.

	01 - Rental Details Rental ID	01 - Rental Details Local Start Time	01 - Rental Details Local End Time	01 - Rental Details Bike ID	01 - Rental Details Duration In Seconds Uncapped	03 - Rental Start Station ID	03 - Rental Start Station Name	02 - Rental End Station ID	02 - Rental End Station Name	User Type	Member Gender	05 - Member Details Member Birthday Year
1063773	23428148	2019-06-28 15:05:05	2019-06-28 15:20:21	3978	916.0	6	Dusable Harbor	18	Wacker Dr & Washington St	Subscriber	Male	1978.00000
827757	23150760	2019-06-13 16:46:26	2019-06-13 17:15:52	1795	1,766.0	251	Clarendon Ave & Leland Ave	35	Streeter Dr & Grand Ave	Subscriber	Female	1989.00000

	ride_id	started_at	ended_at	start_station_id	start_station_name	end_station_id	end_station_name	member_casual
441628	22699127	2019-05-17 08:53:21	2019-05-17 09:06:18	53	Wells St & Huron St	25	Michigan Ave & Pearson St	Customer
458638	22719524	2019-05-18 23:42:00	2019-05-18 23:46:35	300	Broadway & Barry Ave	13	Wilton Ave & Diversey Pkwy	Subscriber
73628	22264727	2019-04-09 11:01:55	2019-04-09 11:31:02	134	Peoria St & Jackson Blvd	85	Michigan Ave & Oak St	Subscriber
1028828	23386204	2019-06-26 17:19:22	2019-06-26 17:32:30	100	Orleans St & Merchandise Mart Plaza	175	Wells St & Polk St	Customer
695001	22995954	2019-06-05 09:48:28	2019-06-05 10:11:59	52	Michigan Ave & Lake St	91	Clinton St & Washington Blvd	Customer

	ride_id	started_at	ended_at	start_station_id	start_station_name	end_station_id	end_station_name	member_casual
1424754	24999563	2019-09-17 08:38:32	2019-09-17 08:50:24	96	Desplaines St & Randolph St	181	LaSalle St & Illinois St	Subscriber
984298	24539744	2019-08-23 07:37:59	2019-08-23 07:52:23	54	Ogden Ave & Chicago Ave	287	Franklin St & Monroe St	Subscriber
1069449	24629068	2019-08-27 19:05:13	2019-08-27 19:40:40	284	Michigan Ave & Jackson Blvd	195	Columbus Dr & Randolph St	Customer
1245296	24812719	2019-09-06 16:17:04	2019-09-06 16:31:36	195	Columbus Dr & Randolph St	338	Calumet Ave & 18th St	Subscriber
93242	23587558	2019-07-06 13:02:53	2019-07-06 13:22:34	324	Stockton Dr & Wrightwood Ave	268	Lake Shore Dr & North Blvd	Subscriber

	ride_id	started_at	ended_at	start_station_id	start_station_name	end_station_id	end_station_name	member_casual
268260	25502923	2019-10-20 18:46:43	2019-10-20 18:59:37	57	Clinton St & Roosevelt Rd	21	Aberdeen St & Jackson Blvd	Customer
264234	25498730	2019-10-20 13:50:14	2019-10-20 13:55:31	153	Southport Ave & Wellington Ave	226	Racine Ave & Belmont Ave	Subscriber
577900	25830311	2019-12-05 14:30:50	2019-12-05 14:35:23	620	Orleans St & Chestnut St (NEXT Apts)	359	Larrabee St & Division St	Subscriber
486252	25732882	2019-11-19 13:55:22	2019-11-19 14:06:03	98	LaSalle St & Washington St	71	Morgan St & Lake St	Subscriber
18874	25243406	2019-10-02 06:16:05	2019-10-02 06:53:34	324	Stockton Dr & Wrightwood Ave	375	Sacramento Blvd & Franklin Blvd	Subscriber

	ride_id	started_at	ended_at	start_station_name	start_station_id	end_station_name	end_station_id	member_casual
379338	0376DF07868FADA6	2020-03-23 19:10:00	2020-03-23 19:20:16	Larrabee St & Division St	359	Wells St & Hubbard St	212.00000	member
101024	1E22A26BBF936DAF	2020-01-05 13:55:57	2020-01-05 14:26:26	Streeter Dr & Grand Ave	35	Theater on the Lake	177.00000	casual
14176	B56BC954825AB21D	2020-01-26 21:24:19	2020-01-26 21:31:02	Dearborn St & Monroe St	49	Dearborn St & Erie St	110.00000	member
298096	277E70C8F5E5E001	2020-03-12 06:53:50	2020-03-12 07:00:34	Canal St & Adams St	192	Michigan Ave & Madison St	197.00000	member
98637	C18C6AB988A2F753	2020-01-03 17:42:08	2020-01-03 17:44:54	Wacker Dr & Washington St	18	Clinton St & Lake St	66.00000	member

Table of Contents

Introduction¶

Preparing Data¶

Processing data¶

Overview¶

Duplicated rows¶

Missing values¶

Data types¶

Adding new columns¶

Ride duration¶

Weekday¶

Month¶

Hour¶

Trimming outliers¶

Data integrity¶

Analysis¶

Month¶

Weekday¶

Hour¶

Workday hours¶

Weekend Hours¶

Ride duration¶

Conclusion, Recommendations¶

	ride_id	started_at	ended_at	start_station_id	start_station_name	end_station_id	end_station_name	member_casual
273376	F55F68EA8D8AAF03	2020-02-19 16:20:53	2020-02-19 16:24:12	283	LaSalle St & Jackson Blvd	284.00000	Michigan Ave & Jackson Blvd	member
308979	D4E7959A67DB44E6	2020-03-25 12:43:34	2020-03-25 12:50:01	36	Franklin St & Jackson Blvd	321.00000	Wabash Ave & 9th St	member
253257	03FC48C474DEAD82	2020-02-02 15:55:33	2020-02-02 15:57:36	59	Wabash Ave & Roosevelt Rd	321.00000	Wabash Ave & 9th St	member
193379	714D92EB0145A71A	2020-02-25 09:17:48	2020-02-25 09:21:23	283	LaSalle St & Jackson Blvd	81.00000	Daley Center Plaza	member
118126	A70D7DE47E92DC27	2020-01-07 14:43:35	2020-01-07 14:51:31	423	University Ave & 57th St	419.00000	Lake Park Ave & 53rd St	member

	ride_id	started_at	ended_at	start_station_id	start_station_name	end_station_id	end_station_name	member_casual
1508604	25086316	2019-09-21 15:52:50	2019-09-21 15:58:05	199	Wabash Ave & Grand Ave	31.00000	Franklin St & Chicago Ave (Temp)	Subscriber
5984	23486723	2019-07-01 12:18:42	2019-07-01 12:45:41	623	Michigan Ave & 8th St	35.00000	Streeter Dr & Grand Ave	Customer
71467	25298452	2019-10-06 10:59:27	2019-10-06 11:05:34	111	Sedgwick St & Huron St	110.00000	Dearborn St & Erie St	Customer
672746	22969870	2019-06-03 18:27:34	2019-06-03 18:31:32	260	Kedzie Ave & Milwaukee Ave	501.00000	Richmond St & Diversey Ave	Subscriber
190985	23704301	2019-07-12 07:36:11	2019-07-12 08:00:57	334	Lake Shore Dr & Belmont Ave	464.00000	Damen Ave & Foster Ave	Customer

	month	is_member	count
0	Jan	casual riders	7592
1	Jan	members	135057
2	Feb	casual riders	12110
3	Feb	members	125541
4	Mar	casual riders	24335

	weekday	is_member	count
0	Mon	casual riders	102880
1	Mon	members	471519
2	Tue	casual riders	90037
3	Tue	members	507748
4	Wed	casual riders	91978

	hour	is_member	duration_min	count
0	0	casual riders	22	3955
1	0	members	9	7927
2	1	casual riders	23	2364
3	1	members	9	3969
4	2	casual riders	21	1324

	hour	is_member	count
0	0	casual riders	4308
1	0	members	7777
2	1	casual riders	3056
3	1	members	4966
4	2	casual riders	1980